A Retrospective Analysis based on Multiple Machine Learning Models to Predict Lymph Node Metastasis in Early Gastric Cancer (EGC)
EAES Academy. Yang T. 07/05/22; 366561; P304
CLICK HERE TO LOGIN
REGULAR CONTENT
REGULAR CONTENT
Login now to access Regular content available to all registered users.
Abstract
Discussion Forum (0)
Rate & Comment (0)
Background: Endoscopic submucosal dissection has become the primary option of treatment for early gastric cancer (EGC). However, lymph node metastasis may lead the prognosis. We analyzed factors related to lymph node metastasis in this group of patients and we developed a construction prediction model with machine learning using data from a retrospective series.
Methods: Two independent cohorts series were evaluated including 305 patients with EGC from China as training set, and 35 patients from Spain as a validation set. Five classifiers obtained from machine learning were selected to establish a robust prediction model for lymph node metastasis in EGC.
Results: The clinical variables such as invasion depth, histologic type, ulceration, tumor location, tumor size, Lauren classification, and age were selected to establish the five prediction models. Linear Support Vector Classifier (Linear SVC), Logistic regression model, Extreme Gradient Boosting model (XGBoost), Light gradient boosting machine model (LightGBM), and Gaussian Process Classification model. Interestingly, all prediction models of training set showed an accuracy between 70% and 81%. Furthermore, the prediction models of the validation cohort exhibited an accuracy between 48% and 82%. The Areas Under Curve (AUC) of the five models between the training and validation group were between 0.736 and 0.830.
Conclusions: Our results support that machine learning methods could be used to predict lymph node metastasis in early gastric cancer and improve patient management accordingly.
Methods: Two independent cohorts series were evaluated including 305 patients with EGC from China as training set, and 35 patients from Spain as a validation set. Five classifiers obtained from machine learning were selected to establish a robust prediction model for lymph node metastasis in EGC.
Results: The clinical variables such as invasion depth, histologic type, ulceration, tumor location, tumor size, Lauren classification, and age were selected to establish the five prediction models. Linear Support Vector Classifier (Linear SVC), Logistic regression model, Extreme Gradient Boosting model (XGBoost), Light gradient boosting machine model (LightGBM), and Gaussian Process Classification model. Interestingly, all prediction models of training set showed an accuracy between 70% and 81%. Furthermore, the prediction models of the validation cohort exhibited an accuracy between 48% and 82%. The Areas Under Curve (AUC) of the five models between the training and validation group were between 0.736 and 0.830.
Conclusions: Our results support that machine learning methods could be used to predict lymph node metastasis in early gastric cancer and improve patient management accordingly.
Background: Endoscopic submucosal dissection has become the primary option of treatment for early gastric cancer (EGC). However, lymph node metastasis may lead the prognosis. We analyzed factors related to lymph node metastasis in this group of patients and we developed a construction prediction model with machine learning using data from a retrospective series.
Methods: Two independent cohorts series were evaluated including 305 patients with EGC from China as training set, and 35 patients from Spain as a validation set. Five classifiers obtained from machine learning were selected to establish a robust prediction model for lymph node metastasis in EGC.
Results: The clinical variables such as invasion depth, histologic type, ulceration, tumor location, tumor size, Lauren classification, and age were selected to establish the five prediction models. Linear Support Vector Classifier (Linear SVC), Logistic regression model, Extreme Gradient Boosting model (XGBoost), Light gradient boosting machine model (LightGBM), and Gaussian Process Classification model. Interestingly, all prediction models of training set showed an accuracy between 70% and 81%. Furthermore, the prediction models of the validation cohort exhibited an accuracy between 48% and 82%. The Areas Under Curve (AUC) of the five models between the training and validation group were between 0.736 and 0.830.
Conclusions: Our results support that machine learning methods could be used to predict lymph node metastasis in early gastric cancer and improve patient management accordingly.
Methods: Two independent cohorts series were evaluated including 305 patients with EGC from China as training set, and 35 patients from Spain as a validation set. Five classifiers obtained from machine learning were selected to establish a robust prediction model for lymph node metastasis in EGC.
Results: The clinical variables such as invasion depth, histologic type, ulceration, tumor location, tumor size, Lauren classification, and age were selected to establish the five prediction models. Linear Support Vector Classifier (Linear SVC), Logistic regression model, Extreme Gradient Boosting model (XGBoost), Light gradient boosting machine model (LightGBM), and Gaussian Process Classification model. Interestingly, all prediction models of training set showed an accuracy between 70% and 81%. Furthermore, the prediction models of the validation cohort exhibited an accuracy between 48% and 82%. The Areas Under Curve (AUC) of the five models between the training and validation group were between 0.736 and 0.830.
Conclusions: Our results support that machine learning methods could be used to predict lymph node metastasis in early gastric cancer and improve patient management accordingly.
Code of conduct/disclaimer available in General Terms & Conditions
{{ help_message }}
{{filter}}